* The Kubernetes API server now supports the ability to break large LIST calls into multiple smaller chunks.  A client can specify a limit to the number of results to return, and if more results exist a token will be returned that allows the client to continue the previous list call repeatedly until all results are retrieved.  The resulting list is identical to a list call that does not perform chunking thanks to capabilities provided by etcd3.  This allows the server to...

  }

  private static final int DISTINCT_CHARS = Character.MAX_VALUE - Character.MIN_VALUE + 1;

  /**
   * This is the actual implementation of {@link #precomputed}, but we bounce calls through a method
   * on {@link Platform} so that we can have different behavior in GWT.
   *
   * <p>This implementation tries to be smart in a number of ways. It recognizes cases where the

- Switch to VM Update call in attach/detach disk operation, original CreateOrUpdate call may lead to orphaned VMs or blocked resources ([#81208](https://github.com/kubernetes/kubernetes/pull/81208), [@andyzhangx](https://github.com/andyzhangx))

  std::vector<std::pair<string, string>> grads = GetGradDefs(gdef);
  ASSERT_EQ(1, grads.size());
  ASSERT_EQ(func_name_, grads[0].first);
  ASSERT_EQ("MyGrad", grads[0].second);

  // These calls must be noops
  TF_GraphCopyFunction(host_graph_, func_, grad_func, s_);
  ASSERT_EQ(TF_OK, TF_GetCode(s_)) << TF_Message(s_);
  TF_GraphCopyFunction(host_graph_, func_, nullptr, s_);

- Avoid GCE API calls when initializing GCE CloudProvider for Kubelets. ([#90218](https://github.com/kubernetes/kubernetes/pull/90218), [@wojtek-t](https://github.com/wojtek-t)) [SIG Cloud Provider and Scalability]

* GCE provider: Create TargetPool with 200 instances, then update with rest ([#27865](https://github.com/kubernetes/kubernetes/pull/27865), [@zmerlynn](https://github.com/zmerlynn))
* GCE provider: Limit Filter calls to regexps rather than large blobs ([#27741](https://github.com/kubernetes/kubernetes/pull/27741), [@zmerlynn](https://github.com/zmerlynn))

	madminSysConfig        = grid.NewJSONPool[madmin.SysConfig]()
	madminSysErrors        = grid.NewJSONPool[madmin.SysErrors]()
	madminSysServices      = grid.NewJSONPool[madmin.SysServices]()

	// Request -> Response RPC calls
	deleteBucketMetadataRPC        = grid.NewSingleHandler[*grid.MSS, grid.NoPayload](grid.HandlerDeleteBucketMetadata, grid.NewMSS, grid.NewNoPayload).IgnoreNilConn()

#### Metrics
* [stable] Add support for cloudprovider metrics for storage API calls. See [Controller manager metrics](https://kubernetes.io/docs/concepts/cluster-administration/controller-metrics/) for more information.

### **Other notable changes**

#### Admission plugin

## Changes by Kind

### Bug or Regression

- Kube-apiserver: gzip compression switched from level 4 to level 1 to improve large list call latencies in exchange for higher network bandwidth usage (10-50% higher). This increases the headroom before very large unpaged list calls exceed request timeout limits. ([#112401](https://github.com/kubernetes/kubernetes/pull/112401), [@shyamjvs](https://github.com/shyamjvs)) [SIG API Machinery]

    assertNotSame(value, map.get(key, loader));
    assertEquals(1, loader.getCount());
    assertEquals(1, segment.count);
  }

  @AndroidIncompatible // Perhaps emulator clock does not update between the two get() calls?
  public void testComputeExpiredEntry() throws ExecutionException {
    CacheBuilder<Object, Object> builder = createCacheBuilder().expireAfterWrite(1, NANOSECONDS);
    CountingLoader loader = new CountingLoader();